1. The Field of the Invention
The present invention relates generally to extended drain field effect transistors and electrical power protection circuits. More specifically, the present invention relates to double-sided extended drain field effect transistors and an integrated overvoltage and reverse voltage protection circuit that uses double-sided extended drain field effect transistors.
2. Background and Related Art
Electrical circuits are in widespread use with a practically limitless variety of applications. Innovation in circuit design has changed the very way we live and work. Nevertheless, there are limits to electrical circuitry. One limit is that circuitry is only designed to operate with certain applied electrical voltages. If those supplied electrical voltages should vary from the designed electrical voltage conditions, circuitry may be damaged or destroyed. Such damage or destruction is most often undesired and may often even be unacceptable. Accordingly, various electrical protection circuits have been developed.
The purpose of such protection circuits is to ensure that a load electrical circuit is protected from anomalous applied electrical conditions. Some protection circuits substantially stop any power at all from being delivered to the load circuit if anomalous voltage conditions are applied to the protection circuit. Other protection circuits called voltage regulators control the supplied voltage such that useful voltages are applied to the load circuit despite anomalous supply voltages being externally applied. Voltage regulators are especially useful when it is necessary for the load circuit to continue to operate despite excessive applied supply voltages.
One application in which anomalous applied voltages may occur is in the automotive environment in which a battery supplies power to circuitry throughout the vehicle. In many cases, the vehicle circuitry is not powered directly from the battery. Instead, the vehicle circuitry is powered by a regulator that lowers the battery voltage and smoothes out the transients in the battery voltage. For example, a typical regulator may receive a battery voltage of up to approximately 16 volts and use that voltage to generate a regulator voltage of just 5 volts.
It is not usual in such a vehicle to have the vehicle circuitry experience transient surges of excessive voltages. Protection from such excessive voltages is often termed “overvoltage protection.” Excessive voltage may occur in a vehicle, for example, when certain vehicle circuitry, which is supposed to be supplied by a lower regulated voltage, is directly connected to a battery, which characteristically supplies a much higher voltage. Excessive voltage may also occur in what is called a “double-battery condition” in which two batteries are wrongly connected in series instead of in parallel during a jump-start. This double-battery condition can raise the supplied voltage up to the range from 25 to 30 volts. Yet another type of overvoltage is called “load dump” which occurs when the load current in a vehicle alternator is interrupted. Voltage peaks in the range of 40 to 80V can be generated under the load dump condition.
Also, sometimes the battery may be inadvertently connected in reverse, in which case the load circuitry may be subject to applied electrical voltages that have opposite polarity as compared to the designed applied voltages. Without protection, this may result in normally reverse-biased PN semiconductor junctions to be forward-biased, which can devastate semiconductor-based circuitry. Protection from such reverse power voltages is often called “reverse voltage protection.” Overvoltage and reverse voltage conditions, while common in automotive applications, may occur with significant regularity in other applications as well.
Accordingly, protection circuits have been developed to act as a buffer between an applied supply voltage and the load circuitry such that when anomalous supply voltages occur, the anomalous supply voltage is either blocked from reaching the load circuitry entirely, or else the anomalous supply voltage is regulated such that the voltage applied to the load circuitry is appropriate. It is advantageous to have such protection circuits be integrated (e.g., on the same semiconductor die) as the load circuit being protected. This reduces the cost, size and power requirements of the combination of the protection and load circuits.
Some protection circuits are designed for reverse voltage protection, while others are designed for overvoltage protection. Integrating such functionality into a single circuit would result in a smaller design than a combination of non-integrated reverse voltage and overvoltage protection circuits. The less complex the design, the less area is consumed on the chip. With increasing amounts of functionality being incorporated onto a single chip, and with chips being incorporated into increasingly confined areas, it becomes increasingly important to minimize as much as is reasonable the amount of room any one circuit on the chip occupies.
In addition, many protection circuits use depletion mode transistors (normally on) or devices such as bipolar transistors, Schottky diodes, or other types of devices that also may not be available in many standard Complementary Metal-Oxide-Silicon (CMOS) processes. Even if available, the process complexity involved with fabricating such devices is higher. Thus, the presence of such devices increases the cost of the circuit. Also, such devices may significantly increase the voltage drop across the protection circuit even if the applied supply voltages are within the designed tolerances of the load circuit. Many protection circuits also may require oscillators, charge pumps, DC-DC converters, or other complex circuits that may significantly increase the cost, size and power dissipation of the protection circuit.
Therefore, what is desired are overvoltage protection circuits that may have integrated reverse voltage protection functionality incorporated therein, and which may be suitable for integration with the load circuit being protected. It would also be advantageous if the protection circuit had a smaller design that did not require complex circuitry or devices that are difficult to fabricate using standard CMOS processes.
Typical voltage regulators and reverse voltage protection circuits do not use doubled-sided extended drain field effect transistors. Instead, conventional extended drain field effect transistors are used for switching of higher voltages. One conventional extended drain field effect transistor that may be fabricated using standard CMOS processes is illustrated in cross-section in FIG. 4 as transistor 400.
The transistor 400 is a p-channel transistor that is fabricated in an n-well 401 within a p-type substrate 411. The transistor 400 further includes a gate terminal 402, a source terminal 404, and a drain terminal 407. Field oxides regions 403A, 403B, and 403C are positioned as illustrated. Terminal 405 is used to bias the n-well 401. Oxide layer 410 represented by regions 410A, 4101B, 410C and 410D overlies portions of the transistor 400 to provide protection and electrical isolation from an upper metal layer except through the via holes represented by the gaps in the oxide layer 410.
Unlike conventional field-effect transistors, the gate terminal 402 of the extended drain field effect transistor is laterally separated from the drain terminal 407. In particular, a more lightly doped p-region 408 (also referred to as a p-RESURF region) is implanted between the gate terminal 402 and the drain terminal 407. “RESURF” is short for “REduced SURface Field”. The p-RESURF 408 operates to electrically connect the drain terminal 407 to the channel region underneath the gate terminal 402. The p-RESURF region 408 also serves as a region that may sustain large voltage drops in cases when the voltage at the channel region far exceeds the voltage at the drain terminal 407. Accordingly, large differential voltages may be isolated from the remaining circuitry even: when the transistor is switching high voltages.
An additional p-region 409 is laterally disposed on the other side of the p+ drain terminal 407. Together, the p-RESURF regions 408 and 409 significantly increase the breakdown voltage of the p+ drain terminal 407 with respect to the n-well 401. Accordingly, the transistor 400 is well suited for switching high voltages.
The extended drain transistor 400 thus has increased breakdown voltage at the drain. This is sufficient for switching high voltages. However, in other applications in which the extended drain transistor has not been conventionally incorporated, there may be instances in which increased breakdown voltage for the source terminal would also be advantageous.